Implementation of an Emotional Virtual Creature with a Growth Function Model

In this study, we developed an emotional virtual creature that grows according to a growth function model influenced by user input. Our preceding study proposed a growth model for emotions based on changes in the network structure of a self-organizing map. We applied a multilayer perceptron neural network to generate more sophisticated emotional expressions using growth functions. This model generated behavior similar to affective change, as described in behavioral genetics. Based on the growth model for emotions, we propose a virtual creature with which a user can interact using a tablet computer. This creature resembles a bouncing ball and changes color depending on its emotion. The user can pet or tap the creature, play with it using some items, and provide it food. The creature’s actions and emotional expressions are influenced by a various stimuli from the user. We implemented the virtual creature using an Apple iPad and confirmed that the creature grew and learned various emotional expressions through user interaction.